Ensemble model aggregation using a computationally lightweight machine-learning model to forecast ocean waves

Abstract This study investigated an approach to improve the accuracy of computationally lightweight surrogate models by updating forecasts based on historical accuracy relative to sparse observation data. Using a lightweight, ocean-wave forecasting model, we created a large number of model ensembles, with perturbed inputs, for a two-year study period. Forecasts were aggregated using a machine-learning algorithm that combined forecasts from multiple, independent models into a single “best-estimate” prediction of the true state. The framework was applied to a case-study site in Monterey Bay, California. A learning-aggregation technique used historical observations and model forecasts to calculate a weight for each ensemble member. Weighted ensemble predictions were compared to measured wave conditions to evaluate performance against present state-of-the-art. Finally, we discussed how this framework, which integrates ensemble aggregations and surrogate models, can be used to improve forecasting systems and further enable scientific process studies.

[1]  Vivien Mallet,et al.  Ozone ensemble forecast with machine learning algorithms , 2009 .

[2]  Stefan M. Wild,et al.  Bayesian Calibration and Uncertainty Analysis for Computationally Expensive Models Using Optimization and Radial Basis Function Approximation , 2008 .

[3]  Richard J. Beckman,et al.  A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output From a Computer Code , 2000, Technometrics.

[4]  Bruno Sudret,et al.  Efficient computation of global sensitivity indices using sparse polynomial chaos expansions , 2010, Reliab. Eng. Syst. Saf..

[5]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[6]  E. Epstein,et al.  Stochastic dynamic prediction1 , 1969 .

[7]  Jesse D. Roberts,et al.  Numerical modeling of the effects of wave energy converter characteristics on nearshore wave conditions , 2016 .

[8]  Ryan D. Torn,et al.  Boundary Conditions for Limited-Area Ensemble Kalman Filters , 2006 .

[9]  S. James,et al.  A machine learning framework to forecast wave conditions , 2017, Coastal Engineering.

[10]  Peter C. Young,et al.  State Dependent Parameter metamodelling and sensitivity analysis , 2007, Comput. Phys. Commun..

[11]  Scott C. James,et al.  An integrated framework that combines machine learning and numerical models to improve wave-condition forecasts , 2018, Journal of Marine Systems.

[12]  G. Evensen.  Data Assimilation: The Ensemble Kalman Filter , 2006 .

[13]  Mitchell J. Small,et al.  Evaluating response surface designs for uncertainty analysis and prescriptive applications of a large-scale water quality model , 2006 .

[14]  Geir Evensen,et al.  The Ensemble Kalman Filter: theoretical formulation and practical implementation , 2003 .

[15]  J. Xu,et al.  Local wave climate and long-term bed shear stress characteristics in Monterey Bay, CA , 1999 .

[16]  Micha Werner,et al.  Reduction of Monte-Carlo simulation runs for uncertainty estimation in hydrological modelling , 2003 .

[17]  David M. Fratantoni,et al.  Development, implementation and evaluation of a data-assimilative ocean forecasting system off the central California coast , 2009 .

[18]  Ken Perlin,et al.  Accelerating Eulerian Fluid Simulation With Convolutional Networks , 2016, ICML.

[19]  Brendan J. Meade,et al.  Enabling large‐scale viscoelastic calculations via neural network acceleration , 2017, 1701.08884.

[20]  Adrian E. Raftery,et al.  Use and communication of probabilistic forecasts , 2014, Stat. Anal. Data Min..

[21]  C. Leith.  Theoretical Skill of Monte Carlo Forecasts , 1974 .

[22]  B. Iooss,et al.  Global sensitivity analysis for a numerical model of radionuclide migration from the RRC “Kurchatov Institute” radwaste disposal site , 2008 .

[23]  Anuj Karpatne,et al.  Physics-guided Neural Networks (PGNN): An Application in Lake Temperature Modeling , 2017, ArXiv.

[24]  A. Marrel,et al.  Development of a surrogate model and sensitivity analysis for spatio-temporal numerical simulators , 2015, Stochastic Environmental Research and Risk Assessment.

[25]  Enrico Zio,et al.  Polynomial chaos expansion for global sensitivity analysis applied to a model of radionuclide migration in a randomly heterogeneous aquifer , 2013, Stochastic Environmental Research and Risk Assessment.

[26]  Raghavan Srinivasan,et al.  Approximating SWAT Model Using Artificial Neural Network and Support Vector Machine 1 , 2009 .

[27]  Bryan A. Tolson,et al.  Review of surrogate modeling in water resources , 2012 .

[28]  Runze Li,et al.  Design and Modeling for Computer Experiments , 2005 .

[29]  Timothy W. Simpson,et al.  Design and Analysis of Computer Experiments in Multidisciplinary Design Optimization: A Review of How Far We Have Come - Or Not , 2008 .

[30]  Ernesto Arandia,et al.  Surrogate modeling and risk-based analysis for solute transport simulations , 2018, Stochastic Environmental Research and Risk Assessment.

[31]  Dimitri Solomatine,et al.  A novel approach to parameter uncertainty analysis of hydrological models using neural networks , 2009 .

[32]  J. Bidlot,et al.  Intercomparison of the Performance of Operational Ocean Wave Forecasting Systems with Buoy Data , 2002 .

[33]  Purnima Bholowalia,et al.  EBK-Means: A Clustering Technique based on Elbow Method and K-Means in WSN , 2014 .

[34]  Tsuyoshi Murata,et al.  {m , 1934, ACML.

[35]  Emanuele Borgonovo,et al.  Model emulation and moment-independent sensitivity analysis: An application to environmental modelling , 2012, Environ. Model. Softw..

[36]  Manfred K. Warmuth,et al.  Exponentiated Gradient Versus Gradient Descent for Linear Predictors , 1997, Inf. Comput..

[37]  Brian J Reich,et al.  Surface Estimation, Variable Selection, and the Nonparametric Oracle Property. , 2011, Statistica Sinica.

[38]  L. Rosenfeld,et al.  Addressing ocean and coastal issues at the West Coast scale through regional ocean observing system collaboration , 2012, 2012 Oceans.

[39]  Peter A. E. M. Janssen,et al.  Progress in ocean wave forecasting , 2008, J. Comput. Phys..

[40]  G. G. Stokes.  "J." , 1890, The New Yale Book of Quotations.

[41]  Jakub Marecek,et al.  Scaling up Deep Learning for PDE-based Models , 2018, ArXiv.

[42]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[43]  Peter R. Oke,et al.  Ensemble member generation for sequential data assimilation , 2008 .

[44]  E. Epstein,et al.  Stochastic dynamic prediction , 1969 .

[45]  Emanuele Ragnoli,et al.  Characterizing observed circulation patterns within a bay using HF radar and numerical model simulations , 2015 .